Here’s One Thing All the Billion-Dollar Unicorns Have in Common

They’re called unicorns — young companies valued at $1 billion or more — because at one point they were rare. Now, they’re much less so.

In the last two weeks the sector has seen several land large investments: Zenefits got a valuation of $4.5 billion, Pinterest hit $11 billion, Jawbone garnered $3 billion.

And now Uber, the ride-hailing service, is reported to be close to raising a $2 billion investment at a valuation that could reach as high as $50 billion, or worth about as much as FedEx.

(Incidentally, the new word for startups valued above $10 billion, according to Re/code reporter Carmel DeAmicis, is decacorn.)

What’s going on here? How can so many startups achieve the coveted billion-dollar valuation status so readily when investors are supposed to be, by nature, conservative and inherently suspicious of risk? And, of course, is this kind of investing nuts?

A survey out Friday from the Silicon Valley law firm Fenwick and West gives us a pretty good clue. The firm advises a lot of these companies on its funding arrangements with investors — which are generally kept secret — and so it has a solid view into the makings of a unicorn. Fenwick analyzed 37 investments in privately held companies valued at $1 billion or more during the 12 months ended March 31.

It turns out that for companies of a certain size, it’s not that hard to get to unicorn status, provided they’re willing to give their investors a lot of assurances that essentially cover their potential losses. The one thing common to every one of these funding deals, the firm says, is that in every case — all 37 of them — investors demanded a “liquidation preference.”

The phrase refers to language often found in an investment contract — and typical to most VC investments — that gives certain investors the right to get paid first ahead of other parties — such as founders or management — in the event the company is sold. If the company sells for a price that is lower than the valuation the investor paid, that investor is the first one in line to receive the proceeds of the sale until they’re made whole. And if the company sells for a higher price, they’re first in line to reap a share of the profit.

What that ultimately means is the investors are taking on very little risk when investing in unicorns, because they stand almost no risk of losing their money if the company goes south.

Of course, these investments are a gamble that investors make to get more money back, either by way of a public offering or selling at a higher price.

And not all IPOs pay off as handsomely as expected. Box, New Relic and Hortonworks all debuted on public markets at valuations lower than they commanded during their private rounds. Only about 20 percent of the time did investors demand protection against that outcome with what’s called a senior liquidation preference: The investors get paid not only before common investors, but also before those holding preferred stock.

An even smaller percentage of investors in those deals — 16 percent — demanded a minimum IPO price that was at least as high as the valuation they paid, while 14 percent demanded additional shares if the IPO price was lower.

Here’s another way to think about it: According to CB Insights, a database that tracks venture capital investments, the top 10 most valuable unicorn companies were worth a combined $122 billion, and had taken on a combined $12 billion in invested capital. With the liquidation preference, the valuations would have to fall by more 90 percent before their investors suffer any financial loss. Add on a senior liquidation preference and the investors could withstand an even greater decline in the valuation of their investment.

In other words, if you have the right terms, believing in unicorns is not so far-fetched.

If you want to read another great take on deal terms and valuations, don’t miss this one by Heidi Roizen, titled “How to Build a Unicorn From Scratch — and Walk Away With Nothing.”

This article originally appeared on Recode.net.